Resources
Toward a human-centric demand response framework: Optimizing energy efficiency and comfort for aging people in the built environment (pre-print version)
This paper shows how a human-centric demand response framework can improve both energy efficiency and the comfort of older adults in residential buildings. Developed within the DEDALUS project, the framework combines IoT technologies, Artificial Intelligence and personalised comfort models to optimise demand response strategies while addressing the specific needs of elderly occupants. The approach is demonstrated through the DEDALUS Italian pilot in Treviso, showcasing its potential to support more inclusive, user-centred and sustainable building energy management.
Cost Minimization in Energy Communities With Multi-Agent Deep Reinforcement Learning and Linear Programming
This paper presents MODREC, a decentralised Community Energy Management System combining multi-agent deep reinforcement learning and linear programming to optimise energy use in residential energy communities. By coordinating PV systems and battery storage under dynamic pricing, the approach enables intelligent load shifting, reduces costs by up to 30%, and supports the use of residential flexibility as a system-level resource. Improving energy autonomy of positive energy districts using multi-agent deep reinforcement learning
This paper presents CoMAD V2G, a multi-agent deep reinforcement learning solution to improve energy autonomy in Positive Energy Districts with shared energy storage and V2G-enabled electric vehicles. By optimising EV charging and discharging while preserving user comfort, the approach reduces grid dependency, lowers household electricity costs by up to 25%, and supports more sustainable community-level energy management. Integration of Cloud Computing and Real Time Simulation Systems for Validating Energy Management Solutions
This paper presents a Cloud-in-the-Loop architecture to validate cloud-based energy management solutions before real-world deployment. Using real-time simulation and hardware-in-the-loop tools, the approach supports the optimisation of storage use in distributed residential energy systems. Data-driven techniques to improve the reliability of low voltage electricity networks through dynamical evaluation of non-technical losses
This paper presents a data-driven methodology to improve the reliability of low-voltage electricity networks. By detecting and assessing the origin of non-technical losses and grid anomalies, the approach supports smarter and more efficient management of increasingly distributed and variable energy systems. Optimizing Electricity Sharing from Generation and Storage Within District-Based Renewable Energy Communities
This paper presents a linear programming-based methodology to optimise energy distribution within renewable energy communities. Validated in a real pilot with 12 dwellings, a shared PV plant and a centralised storage system, the approach improves fairness and efficiency in local energy allocation. IISA 2024 | Methodology for assuring interoperability in demand-response services (pre-print version)
This paper presents a methodology to ensure interoperability in demand response services, enabling seamless data exchange across building systems, devices and applications. It proposes an interoperable architecture based on a shared data model, extending existing ontologies to better represent buildings, users and performance indicators within the DEDALUS framework. Optimal Photovoltaic Energy Allocation in Residential Renewable Energy Communities
The paper explores improved energy-sharing mechanisms within Renewable Energy Communities. Using enhanced linear programming methods validated with real data from a Spanish DSO, it shows how renewable energy allocation can be optimised to reduce surplus generation and increase user benefits across different REC sizes.
Social -Sciences- and- Humanities-Based Profiling of Energy Consumers Towards Increasing Demand Response Engagement
This paper examines the effectiveness of demand response programmes in different European residential contexts by analysing households’ willingness to take part in energy management strategies. Based on questionnaire data from 284 participants in six countries, it identifies different user profiles and their potential for adaptive DR programmes.
Heuristic based federated learning with adaptive hyperparameter tuning for households energy prediction
Hybrid transformer model with liquid neural networks and learnable encodings for buildings’ energy forecasting
The paper presents a hybrid Transformer-based model for building energy forecasting that improves accuracy by combining Liquid Neural Networks with learnable encodings. The approach addresses key limitations of existing models, particularly in capturing nonlinear dependencies, weather influences and temporal dynamics in energy consumption patterns.
Social Factors in P2P Energy Trading Using Hedonic Games
This paper shows how energy communities can improve resilience and flexibility through peer-to-peer trading. Using a hedonic game model that factors in social relationships, it optimises prosumer coalitions and increases energy transactions by 5% compared to other methods. Blockchain Solution for Buildings’ Multi-Energy Flexibility Trading Using Multi-Token Standards
This paper shows how buildings can enhance energy system resilience by leveraging their flexibility in multi-carrier energy networks. Using blockchain and ERC-1155 tokens, it enables buildings to trade heat and electricity in community marketplaces, improving interoperability, reducing transactional overhead, and supporting integrated energy grids.
Evolutionary game for incentivizing social cooperation of prosumers in transactive energy communities
The paper explores how evolutionary game theory can foster cooperation among prosumers in transactive energy communities. It presents a blockchain-based local energy market that rewards collaborative behaviour and discourages self-interested actions, supporting peer-to-peer trading and flexibility. Demand response optimization for smart grid integrated buildings: Review of technology enablers landscape and innovation challenges
This paper presents a comprehensive overview and analysis of state-of-the-art technological advancements in building integration within smart grids, with a focus on their role in demand response. It consolidates knowledge from high-quality sources on key research topics, providing researchers, building owners, and energy stakeholders with insights into the latest developments, trends, and best practices in the field.
Methodological approach for optimizing demand response in building energy management through AI-enhanced comfort-based flexibility models (pre-print version)
This paper presents a methodology to optimise Demand Response in building energy management systems through AI-enhanced comfort-based flexibility models. By combining historical and real-time data, it aligns energy use with individual comfort preferences and demonstrates its application across several DEDALUS pilots. A machine learning-based framework for clustering residential electricity load profiles to enhance demand response programs
This paper presents a machine learning framework for clustering residential electricity load profiles to support more targeted Demand Response programmes. Using data from thousands of London households and integrating explainable AI, it offers a scalable and transparent approach to consumer segmentation. Review of blockchain tokens creation and valuation
This paper examines the creation and valuation of blockchain tokens, reviewing key standards and the main economic factors affecting their value. Using the PRISMA methodology, it provides an overview of tokenisation applications and related regulatory and valuation challenges. 
